Something to think about regarding Real Estate

Quote from Raystonn:

They even state in that article that the correlation breaks down when you look at earlier time periods. So they've basically cherry picked the time period during which it seems to correlate. If the S&P 500 hadn't correlated they would have looked at the Dow, then the Nasdaq, then the Russell, then a Smallcap index, etc. etc. until they found something that correlated for a few years. This is simple random chance. Given enough charts and enough years, a many will correlate due to complete random chance.

In other words, the correlation is completely meaningless.

-Raystonn
More...

I'm sure that to back up your assertion you have done Pearson and/or Spearman tests. Certainly you have the R2 number close at hand, which would make it easy for you to justify your comment that "this is simple random chance"....

I would love to see the data..otherwise I get the feeling that your comment is "completely meaningless"...
 
But Steve, surely your criticism is not to the point, since Raystonn is accusing the authors of optimisation. On that point the R2 statistic would tell us nothing other than how thoroughly the optimisation has been performed :)
 
I understand..Here is my point...In science, we look at claims by by other researchers and we say "Okay then, if your data is correct and you have adhered to standard scientific method, then I should be able to replicate your result here in my lab (home)".

So then we go to the NAHB site and we notice that there is data going back to 1985 (more than 20 years). We import that data to our site and we do the tests (Spearman, Pearson, whatever) and we compare the result to the claims that have been made.

How did the data correlate in the early years (starting from 1985). Was the R2 above 0.05 or below? What happens in the later years? In other words, we duplicate the tests or if we know of better or easier ways to test we do it....We don't just type in our post what we have read on the internet. Why? because IF WE JUST TAKE WHAT WE READ ON THE INTERNET AS TRUTH, WE DON'T REALLY LEARN ANYTHING. IT MIGHT BE CORRECT, OR IT MIGHT BE BULLSHIT...IF HOWEVER WE TRY TO DUPLICATE THESE TESTS AND WE FIND SOMETHING DIFFERENT, WE MIGHT LEARN SOMETHING VERY IMPORTANT.

For instance we might learn that correlation is "rhythmic". that is to say, in the early years, it might well have been that correlation was not there (between the NAHB and the broader market), but that later in the data series the correlation was strong, back and forth so to speak. If that is true (as it is with a lot of data series) it tells you that at some time in the future you can expect the broader market to "catch up" to the NAHB or conversely you might expect the NAHB to "catch up" to the broader market...either way because you went to the trouble of doing the investigatory work (instead of just parroting what you read) you have not only educated yourself a bit, but you might have learned something that helps you to get onto a trade that makes some money...Just a couple of thoughts.
 
Quote from Raystonn:

They even state in that article that the correlation breaks down when you look at earlier time periods. So they've basically cherry picked the time period during which it seems to correlate. If the S&P 500 hadn't correlated they would have looked at the Dow, then the Nasdaq, then the Russell, then a Smallcap index, etc. etc. until they found something that correlated for a few years. This is simple random chance. Given enough charts and enough years, a many will correlate due to complete random chance.

In other words, the correlation is completely meaningless.

-Raystonn
More...

Kind of. From what I've read 40% of the jobs of late have been driven by homebuilding and much of the economy has been carried by construction and real estate. So there is a coupling of late that explains the correlation.

But your observation is correct in a sense, because this is unlikely to continue in the future...
 
Quote from Quark:

I think widespread recognition of deflation in real estate is still a long way off. From a sentiment standpoint, we appear to be just ending the topping process and starting down the other side.

My observation of markets over the years has convinced me that major (as in many, many years long) tops and bottoms take far longer to complete than most people think. Your post is an example of what is often seen around the time a topping process is near an end - a belief that it's already over just as the new downtrend is getting started. In other words, some people who've observed the beginning of the trend change interpret it as a "correction" of shorter duration, rather than a major trend change.

I'd suggest great patience before calling a low in RE. Not that there can't be bounces back up along the way, but I think the trend is down, and it's going to get very, very nasty before it is finally over, and RE values will reach lows that are "unbelievable" today. Think about it, though, who'd have believed you in August 1982 with short term rates at 8% and the DJIA at 777 if you had said the DJIA would reach 11,000 plus in less than 20 years? "Impossible", right?
More...

I just don't think you're going to see a long downtrend in the major metro areas. Yes, hiccups on the way. But, after all, it's supply and demand...
 
Quote from ShoeshineBoy:

But, after all, it's supply and demand...
More...

An economist might believe that (in her moments of supreme self deception) but a trader wouldn't.
Never forget that perception is truth.
 
Quote from kiwi_trader:

An economist might believe that (in her moments of supreme self deception) but a trader wouldn't.
Never forget that perception is truth.
More...

Not sure what you mean?

If you have a lot of buyers in a land-locked location, which is the case with almost every metro area, then it isn't good for a downtrend. Yes, if there's a recession and a increase in home defaults, you could get a temporary downtrend. But imho the trend will still be up over the medium term in any metro area.
 
You must log in or register to reply here.
Back
Top